Anna Genell

BACKGROUND: In epidemiological studies researchers use logistic regression as an analytical tool to study the association of a binary outcome to a set of possible exposures. METHODS: Using a simulation study we illustrate how the analytically derived bias of odds ratios modelling in logistic regression varies as a function of the sample size. RESULTS: Logistic regression overestimates odds ratios in studies with small to moderate samples size. The small sample size induced bias is a systematic one, bias away from null. Regression coefficient estimates shifts away from zero, odds ratios from one. CONCLUSION: If several small studies are pooled without consideration of the bias introduced by the inherent mathematical properties of the logistic regression model, researchers may be mislead to erroneous interpretation of the results.

Epidemiology and outcome of myeloma are mainly reported from large university centers and collaborative groups, and do not represent ‘real-world’ patients. The Swedish Myeloma Registry is a prospective population-based registry documenting characteristics, treatment and outcome in newly diagnosed myeloma, including asymptomatic and localized forms, with the purpose of improving disease management and outcome. This report presents information on patients diagnosed between 2008 and 2015, including data on first-line treatment in patients diagnosed up to 2014, with a follow up until December 2016. We present age-adjusted incidence, patients’ characteristics at baseline, treatment, response, and survival. Baseline data were available with a 97% coverage in 4904 patients (median age 71 years, males 70 years, females 73 years; 72% were 65 years or older), and at 1-year follow up in 3558 patients with symptomatic disease (92% of patients initially reported). The age-adjusted incidence was 6.8 myeloma cases per 100,000 inhabitants per year. Among initially symptomatic patients (n=3988), 77% had osteolytic lesions or compression fractures, 49% had anemia, 18% impaired kidney function, and 13% hypercalcemia. High-dose therapy with autologous stem cell transplantation was given to 77% of patients aged up to 66 years, and to 22% of patients aged 66–70 years. In the study period, 68% received bortezomib, thalidomide, and/or lenalidomide as part of the first-line treatment, rising from 31% in 2008 to 81% in 2014. In active myeloma, the median relative survival of patients aged 65 years or under was 7.7 years, and 3.4 years in patients aged 66 years and over. Patients diagnosed with myeloma in more recent years were associated with significantly higher rates of complete or very good partial remission (P

BACKGROUND: Automatic variable selection methods are usually discouraged in medical research although we believe they might be valuable for studies where subject matter knowledge is limited. Bayesian model averaging may be useful for model selection but only limited attempts to compare it to stepwise regression have been published. We therefore performed a simulation study to compare stepwise regression with Bayesian model averaging. METHODS: We simulated data corresponding to five different data generating processes and thirty different values of the effect size (the parameter estimate divided by its standard error). Each data generating process contained twenty explanatory variables in total and had between zero and two true predictors. Three data generating processes were built of uncorrelated predictor variables while two had a mixture of correlated and uncorrelated variables. We fitted linear regression models to the simulated data. We used Bayesian model averaging and stepwise regression respectively as model selection procedures and compared the estimated selection probabilities. RESULTS: The estimated probability of not selecting a redundant variable was between 0.99 and 1 for Bayesian model averaging while approximately 0.95 for stepwise regression when the redundant variable was not correlated with a true predictor. These probabilities did not depend on the effect size of the true predictor. In the case of correlation between a redundant variable and a true predictor, the probability of not selecting a redundant variable was 0.95 to 1 for Bayesian model averaging while for stepwise regression it was between 0.7 and 0.9, depending on the effect size of the true predictor. The probability of selecting a true predictor increased as the effect size of the true predictor increased and leveled out at between 0.9 and 1 for stepwise regression, while it leveled out at 1 for Bayesian model averaging. CONCLUSIONS: Our simulation study showed that under the given conditions, Bayesian model averaging had a higher probability of not selecting a redundant variable than stepwise regression and had a similar probability of selecting a true predictor. Medical researchers building regression models with limited subject matter knowledge could thus benefit from using Bayesian model averaging.

Solitary plasmacytoma (SP) and plasma cell leukemia (PCL) are uncommon (3-6%) types of plasma cell disease. The risk of progression to symptomatic multiple myeloma (MM) is probably important for the outcome of SP. PCL is rare and has a dismal outcome. In this study, we report on incidence and survival in PCL/SP, and progression to MM in SP, using the prospective observational Swedish Multiple Myeloma Register designed to document all newly diagnosed plasma cell diseases in Sweden since 2008. Both solitary bone plasmacytoma (SBP) (n=124) and extramedullary plasmacytoma (EMP) (n=67) have better overall survival (OS) than MM (n=3549). Progression to MM was higher in SBP than in EMP (35% and 7% at 2 years, respectively), but this did not translate into better survival in EMP. In spite of treatment developments, the OS of primary PCL is still dismal (median of 11 months, 0% at 5 years). Hence, there is a great need for diagnostic and treatment guidelines as well as prospective studies addressing the role for alternative treatment options, such as allogeneic stem cell transplantation and monoclonal antibodies in the treatment of PCL.

The prevalence of multiple myeloma (MM) is increasing in Nordic countries and the rest of the western world. Patients aged ≥75 years at diagnosis constitute an increasing proportion of all MM patients, but are underrepresented in randomized clinical trials. There is an urgent need for studies of the characteristics, treatment and outcome in this cohort. We present data from two nationwide population-based registries of all MM patients diagnosed in Denmark from January 1, 2005 until February 18, 2020, and in Sweden from January 1, 2008 until December 31, 2019, including treatment data for patients diagnosed until 2018 (Denmark) and 2019 (Sweden). In total 4,647 patients were ≥75 years at diagnosis, compared to 7,378 younger patients. Patients ≥75 years, accounting for approximately 40% of all MM patients, are a distinct cohort with more advanced disease at diagnosis, reflected by higher International Staging System (ISS) stage, and a higher proportion have renal failure and anemia. We found a more gradual introduction of modern medications in the older cohort than in the younger, despite simultaneous changes in guidelines. Compared to the cohorts in randomized controlled trials that guide the treatment of non-transplant eligible patients, we found a higher proportion of patients ≥75 years and presenting with ISS III in the real-world populations. Nevertheless, response rates and survival are increasing, indicating that modern treatment regimens are effective and well tolerated also in elderly MM patients in real-world populations.